The Zappos Customer Support Key Performance Indicators.

Software Advice recently authored an article on the key performace indicators the Zappos Customer Service Team incorporates into their self evaluations titled, "A Zappos Lesson in Customer Service Metrics". The article outlines the key metrics any support operations should take into consideration. Here is an excerpt from the article:

Zappos founder Tony Hsieh literally wrote the book on modern customer service. His title, Delivering Happiness, staked claim on The New York Times Best Sellers list for 27 consecutive weeks. And every publication from Forbes to Fast Company has distinguished the e-retailer as a standout in consumer support.

 

Yet few companies heed Zappos’s example.

 

Take a look at your performance metrics. Is time-to-resolution an important indicator? What about call time? While popular, these KPIs are indicative of a very un-Zappos experience. They don’t “drive service into your culture,” as the author ofThe Zappos Experience, Joseph Michelli, explains to me.

 

 

That doesn’t mean Zappos doesn’t use metrics to influence performance and outcomes. Here’s a breakdown of what the e-tailer measures, why, and how you can use these in your business.
 

Measure Total Call Time, Not Time Per Call

Zappos’s longest call on record lasted more than eight hours. The Customer Loyalty Team member (as they’re called) helped the caller find a cycling shoe on another site when they were out of stock. Then they just started to talk. And talk.

 

“It’s more important that we make an emotional connection with the customer, rather than just quickly getting them off the phone,” says Derek Carder, customer loyalty operations manager for the Zappos.

 

Instead of valuing quick time to resolution or processing high call volumes, Zappos looks at the percentage of a time an agent spends on the phone. This metric–personal service level–is a way to “empower the team to utilize their time how they see best promotes customer loyalty,” Carder says.

 

The agent is expected to spend at least 80 percent of their time in customer-facing interactions. It doesn’t matter if that’s one call, or 100. Reps who achieve this target get to spin “the wheel of happiness” to win gift cards and other rewards. Those who fall below the 80 percent line receive coaching.

 

Quantify and Reward Wow Moments

Zappos utilizes relatively little advertising or traditional marketing. They rely and succeed almost exclusively on word of mouth and customer loyalty.

 

“Customer service creates an environment of one-to-one communication. That intimacy creates a special opportunity to build a relationship as opposed to a top of mind impression through advertising,” Michelli said.

 

They do this by measuring four factors on a 100-point scale called the “Happiness Experience Form:”

“Emotion drives people to share with others,” Michelli tells me.

 

Carder gives the example of a customer who called to find shoes for a wedding. The agent asked about her location and found out she had a friend who was married in the same place. That’s creating a PEC, as opposed to just saying “OK, great.” Now for the wow.

 

Zappos didn’t have the shoe in stock, so the agent looked at four other sites. They didn’t have them, either. So, she called a mall, talked to the manager and found a store that had the right pair. She purchased the shoes and personally mailed them to the customer. She received 100 points for that interaction.

 

At the end of the month, management identifies agents with less than a 50-point average on the Happiness Experience Form. Those agents receive extra training. Top performers are rewarded with paid hours off and other incentives.

Similar to talk time, Zappos doesn’t value time to resolution or quantity of interactions when it comes to chat. The agent is expected to create a PEC and follow the same guidelines as a phone conversation. Zappos also monitors “abandonment time,” or periods when an agent has a session open even though the customer already disconnected from the chat.

 

When agents have an idle chat window they can’t respond to a new exchange. Carder said this could be a symptom of “chat avoidance,” which is comparable to an agent not picking up a call so it rolls over to another responder. Zappos measures the abandonment rate in 30-minute increments. If the team experienced a high ratio of idle chats to active session, this would trigger management to coach agents on duty during that time.

 

This strategy zeroes in on the cause of unproductivity in the chat setting–idle chats. When agents aren’t productive, customers wait longer. And the longer they wait, the more apt they are to abandon the session.

 

A recent Cornell University study estimates call center absenteeism averages near six percent per day. That’s a huge amount of productivity lost.

 

Zappos combats this trend with a program they call Panda. Employees receive a point for every day they miss work or come in late. Staff with zero points in a given period receive a varying number of paid hours off. These hours can be accrued and stacked for an entire paid day off, Carder explains.

 

“We’ve seen a huge return on investment in terms of absentees and numbers on deck,” Carder says. Employees can still use their sick and vacation days without affecting their Panda score.

Legendary scholar, author and business consultant Peter Drucker once said, “What’s measured improves.” Success of course hinges on what is measured.

 

Emphasis on call time and time to resolution yields rushed, impersonal interactions. Instead, Zappos has found a way to measure and reward behaviors that impact customer satisfaction most. It’s worked for them. 

CallFire's offers a medley of SMS and and voice solutions to connect more e-tailers to their customer base. CallFire local and toll free numbers can be connected with complex customer support phone tree IVRs. In addition companies can utilize SMS text messages to as a support channel and more optimally respond to their customers. 

To learn more about customer support best practices and solutions, check out the following resources: